4. UQV (University of Alberta)

posted Oct 30, 2010, 6:23 AM by Jeff Ogden   [ updated Jun 9, 2014, 4:04 AM ]
In 1971 the University of Alberta (UQV) in Edmonton, Alberta, Canada became the fourth member of what became the MTS Consortium.

"1967 - 1979: Timesharing Begins" and "1971 - 1984: The Era of MTS" from "Timeline of Computing Services at the University of Alberta" give a few bits of information on adoption of MTS at UQV, but Gerry Gabel, Dale Bendt, and John Stasiuk have provided more detailed recollections below.

Please contribute any information or memories you may have.



From: Dale Bent
Date: April 18, 2011 12:42:03 PM EDT
To: eff Ogden
Cc: "Gerry Gabel
Subject: Re: a message from your (distant past), asking about MTS

Hello Jeff:   Yes, I'll be happy to help.   Thank you for your initiative in compiling this information and keeping the achievements of the "MTS Community" alive and memorable -- as they deserve to be.   I will compile some info and respond, but not very quickly.   Feel free to remind me if you wish.   With best wishes, Dale Bent.

From: Gerry Gabel
Date: April 18, 2011 7:07:32 PM EDT
To: Jeff Ogden , DaleBent
Cc: John Stasiuk, Alan R Davis
Subject: Re: a message from your (distant past), asking about MTS

Hello Jeff and Dale:

I read with interest the discussions related to non-technical matters in the early years of MTS.   I have some time today to relate some of my memories on those discussions.  Dale can correct or add to my recollections (although that was over 40 years ago!).   Unfortunately, one person who would remember much of this was Daryl Webster who passed away in December.   However, I have copied two others - John Stasiuk and Alan Davis - who may also have remembrances from those days.

I have a plaque on my "home office" wall which is a mounted IBM 2314 disk engraved with 4 of my major accomplishments at UQV presented to me when I left in 1977.  MTS is the second one with a date of 1970.  I came to UQV as a programmer in 1967.  The Computing Center was part of the Computing Science Dept. under Dr. Don Scott.  I was attracted to UQV because they had ordered a then state-of-the-art IBM 360/67. I worked in the group responsible for the operating system software and related sub-systems.  Most of the staff worked on trying to get TSS operating successfully, which it did not and they moved on to OS/360.  Meanwhile, I and a few others (eg Daryl Webster, Alan Davis) worked on other systems such as DOS/APL and CP/CMS.  As I recall, the Computing Center shared the operating day between all three systems - awkward to say the least - with CMS and APL running during the day and OS doing batch work over night.

I was a regular attendee at the IBM SHARE conferences and I think that was where I first became aware  of MTS.  I don't recall the exact year but it was probably around 1969.  We were aware that UBC was running MTS so that may be how we learned of it although I do remember meeting Mike Alexander and Don Boettner at SHARE.  UQV got a distribution tape from UM and began testing MTS.   We were running a "trial" MTS along with CP/CMS and DOS/APL (on a virtual machine) for our timesharing needs during the day and OS/360 for traditional batch.

Around that time I became the Manager of the Systems Programming group and started to promote MTS as the single operating system to meet the Universities computing needs.  We had converted APL to run under MTS as well as our Student Oriented Batch Facility (SOBF).  TSS was essentially dead, CP/CMS had some success but was not as efficient.  Users who wanted OS/360 accepted over night batch service primarily for production runs as they could do development under MTS.  Also, the University's Admin Dept got their own OS machine in 1969.  By late 1970, the Dept. of Computing Services was being split from Computing Science and Dale Bent became the first Director in 1971.   At the time, Dale was a regular user of computing services as a Professor in the Business Admin Faculty and a keen supporter of timesharing.

I don't recall any pivotal event or major decision point that resulted in MTS going into full production on Jan.1, 1971.  I never did any presentation to some governing body or computing oversight committee.  I think the decision to adopt MTS just evolved over time as more and more users preferred it to the alternatives.  The University had adopted a fee for service policy in 1970 and MTS had a superior charging system as well.  Adopting MTS finally fulfilled the expectations of the time sharing/virtual memory capabilities of the IBM 260/67 installed almost 4 years earlier.

Perhaps Dale or John Stasiuk or Al Davis can add to this or correct my recollections.  John was a programmer there before me.

I also read the other pages you have links to and thought that John Hogg's musings about why MTS was not more widely adopted was spot on.  In that era, many organizations were reluctant to go away from the mainline IBM operating systems and it was universities or research labs that were prepared to take the risk and enjoy the rewards.  I have often wondered what would have happened if a company such as Amdahl had licensed MTS and offered it on their machines.

Thank you again, Jeff, for undertaking this project to document the life of MTS and to keep the discussions going.   It was a thoroughly enjoyable and rewarding part of my career.   Best regards, Gerry

At 8:40 PM -0600 4/18/11, John Stasiuk wrote:
Hi Dale & Gerry,
 
       Always a pleasure to hear from you guys. Yes the MTS story needs to be recorded so that some people can see how far ahead it really was. I think everyone involved with MTS is proud of their contributions to a great system. Too bad it wasn't able to evolve into an ongoing system.
 
        I also feel much the same way about Textform. It was far ahead of all other text processing systems. Too bad we didn't sell it to Microsoft. Word would have been a much better product, much earlier than it turned out to be. Grant was successful in having much of the Textform system buried into HTML so at least some of the system survived.
 
        I didn't see any errors in Gerry's comments. I feel they reflect what actually happened.
 
        I too was saddened to learn of Daryl's passing. He was a great individual and friend and went far too early. Unfortunately we have no control over this but must work to keep those memories alive. He worked hard to make MTS a better System.
 
I look forward to seeing more comments on MTS.
Best of health to all of you.
 
take care,     John

From: Gerry Gabel
Date: April 19, 2011 12:22:15 AM EDT
To: John Stasiuk
Cc: "Dale Bent", "Alan R Davis", Jeff Ogden
Subject: Re: a message from your (distant past), asking about MTS

Greetings to you all:

Thanks for reviewing what I passed on to Jeff.  I realize our memories grow dim after 40+ years but what memories do remain are good ones.

Dale - I look forward to your comments.   You also have the perspective of an external user of the variety of services available before becoming an "insider" as the Director of CS.

Jeff: there is more to come so best you wait until you have all our input into your valuable record.

Best regards,  Gerry

From: Dale Bent
Date: June 3, 2011 4:37:20 PM EDT
To: Jeff Ogden, Gerry Gabel
Subject: Notes about MTS at the University of Alberta

Hello Jeff

Thanks for tracking me down at Athabasca University, where I am still teaching MBA classes online.  AU operates the first and most successful distance learning network in Canada.   However, you never know when I may retire, or "pop off" as they say in the Alumni Office.   A more reliable long-term way of reaching me or my survivors would be at my home email.

I got my start with computers about 1958 when the first computer at the University and about the third in Alberta arrived on campus!   The Physics Department had bought an LGP-30 magnetic-drum memory computer with 4096 31-bit words of memory and a cycle time of 17.5 milliseconds, lightning-fast by chalk-board standards.    After hanging around the computer centre for a few years, and going away to get my Ph.D. at Stanford, I came back to campus as an Assoc. Prof. of Business.   I always did a lot of computer work, including Statistical Analysis using the SPSS statistical package, which I and a friend invented while at Stanford.

The University of Alberta Computer Centre had an IBM 360/67, based on the forlorn hope that a virtual-memory operating system would become available from IBM.   It didn't appear, and the computer centre was exploring (and using!) various alternatives.   OS/360 with variable numbers of partitions, a special student application (SOB) continuously running, CP/CMS, and MTS were tried.   They had kicked tires on other alternatives also. The computer centre was being operated by the Computer Science Department, based on the whimsical idea that they would be best qualified to do it. They knew about as much about operating a university service department as present-day computer science grads know about business.

In 1969, I was just another complaining user of the Computer Centre.  I had become an active member of a campus vigilante group known as "CURC" -- pronounced "curse" -- the Computer User's Representatives Committee. Having come from Stanford, which arguably had the world's best computer
centre at that time,  I was plenty frustrated and angry at the limited on-line offerings such as "APL 10-11am, M,W,F" which seldom came up on time, and frequently crashed.  Under fire from CURC and others, Bill Adams, the Assoc. Director of the Computer Centre, and Assoc. Professor of Computer Science, resigned.

One day I was sitting at my desk in the Business School, probably waiting for the time-sharing to come on-line, when a person from the Computer Centre who several from the early days would remember,  Vic Yanda. called. Vic said "Dr. Bent, you've been quite critical of us over here, haven't you? "   I said "Yes, that's true."  He said "Well if you're so goddam smart why don't you come over here and run things?", and hung up the phone. That was typical Vic Yanda, but thinking it over, I decided that it might be fun to come over for a few months, straighten things out, and then resume my serious career as a Professor of Management Science.

So I became the Assoc. Director of the Computer Centre.  My first priority was to implement one operating system that provided good quality time-sharing that could be relied upon, 24-7.  Personally, I didn't much care what it was, provided it met most needs, and could reliably run continuously.  Gerry Gabel was the Manager of the Systems Group, and most knowledgeable concerning the Operating Systems alternatives available to us.   Based on his knowledge of the development work for MTS going on at Michigan, and experience at University of British Columbia, Gerry recommended going with MTS.   In retrospect, I guess both Gerry and I "bet our careers" on this choice, although I was now the one who had to face the monthly meeting with CURC.

Although I was young and naive, I did recognize that the best time to implement our new operating schedule was during the summer, while many academic are away from campus, and there were few students around.  We briefly announced our plans, and bang! -- MTS it was, 24-7.  I fielded the hate mail shoved under my door by the most serious users on campus -- professors who has mastered the arcane art of using IBM's JCL (Job Control Language).  Gerry and his capable staff worked hard to stabilize MTS operations for our heavy fall/winter usage.

Our pre-emptive strike worked.  A separate Department of Computer Services was established in the following year, and I became the first Director. By the time serious opposition to MTS developed, many users had come to enjoy the clean, user-friendly functions, enough to overwhelm the few persons whose sophisticated JCL card-decks were now only archaelogical curiosities. In fact, we worked hard to help serious (as opposed to abusive) users who had serious work to do, and many of them became our most ardent supporters.

It became more and more clear that our decision to implement MTS was right for the University of Alberta.   In addition to offering effective technology making good use of our IBM/67 hardware, we had joined together
with several very competent computing centres, all devoted to first class academic computing.  We certainly benefited from the deep "systems programming" knowledge at the other MTS computer centres, but our unique contributions to the MTS community resulted from adapting excellent application software to the needs of MTS users.  Notable were the implementations we made of SPSS and other statistical software, SPIRES -- a data-based management system developed at Stanford, the APL processors and application library, and Textform.

Today it almost seems comical that a text-processor could have been main contribution of a mainframe computer service.   (Later, Word Perfect, and something called Word became popular.)   But at the time, Textform became the writing workhorse of the academic community at Alberta.   Grant Crawford, who later became the Associate Director, and his team wrote software to drive Xerox page printers from the Textform processor.  These page printers ran night and day, and were used by everyone on campus.   It was our "killer app".    We had a delivery service which delivered physical computer output to every building on campus.  Faculty with a terminal could turn around a job in an hour or two without freezing walking back and forth
to the computing centre (Very important in the "Great Canadian North").

Powered by MTS, the Department of Communications and Computer Services had grown to a unit of about 120 persons.  In cooperation with the other MTS universities, we implemented a large time-sharing network and many, many classrooms, all offering MTS.  Rolly Noel and others from Gerry Gabel's systems group continued to develop the network offerings.  A lot of our growth was paid by offering services off-campus.  At its peak, about $3 annually were received from off-campus customers and this was used to make on-campus computer work less and less expensive.  Engineering and other commercial companies came to the University because nothing camparable in ease of use and price was available locally.  The annual computing budget,
capital and operating peaked at about $25 million before 1985, when I left Alberta to go to the University of Western Ontario.

Thank you for the opportunity to offer this sketchy overview.  Lots more happened, but hopefully the above notes accurately show the huge effect the MTS operating system had upon the University of Alberta computing.

I never did get back to a respectable academic career.  I was having too much fun.

Please contact me if I can clarify or correct any of the above.   I'm looking forward to reading more online.   And thanks to you, Jeff, for getting in touch and offering this opportunity for sharing my fond memories
of the glory days of MTS.

SIncerely, Dale Bent.

At 5:08 PM -0400 6/3/11, Jeff Ogden wrote:
Thanks Dale. It was fun reading.

Can you say a little bit more about  Vic Yanda?  What was his role at the CC? Was he trying to get you to come over or was he just letting off steam when he called?

I assume the $3 is really $3M from external users.

You were the director when UofA purchased an Amdahl 470v/6 (P5), the first Amdahl sold in Canada, if I've got the story right?  Can you say anything about that switch?  Again, if I remember correctly, it started a long relationship with Amdahl.  I assume that it didn't make IBM happy and it probably made some administrators at UofA uneasy.  Was IBM already out of sorts because you were running MTS or were they just happy that you kept the 360/67 when TSS wasn't delivered?

  -Jeff

From: Gerry Gabel
Date: June 3, 2011 6:59:17 PM EDT
To: Jeff Ogden
Cc: dale bent
Subject: Re: Notes about MTS at the University of Alberta

Jeff and Dale:

First, Dale, thanks for that excellent historical memoir especially the part about Vic challenging you to fix things at the Computing Centre.  I did not know that.   We do have to get together to exchange more memories.   I'm not on the Tigh-Na-Mara Strata Council anymore so I don't get up your way too often.  So if you're down in Victoria, please give us a call.

Jeff:  I can answer some of your questions.  Yes, UVQ installed the first Amdahl in Canada - a 470V/6, serial number 5.  We selected it after doing an extensive benchmark test that involved simulating a multi user online script.  We did this at IBM Poughkeepsie and at Amdahl in Sunnyvale.  One of the photos I sent you was from the benchmarking trip; both Dale and I and others where there.  I think IBM lobbied hard to keep Amdahl out but in our assessment, Amdahl had the superior machine.  I don't think they resented us running MTS, as I recall one of the IBM SEs actually helped us develop code.   The sales people just didn't want any other brand but IBM in the Centre.  They probably met with the senior University administration but we had good support from them for what we wanted to do. Later the machine was upgraded to a V/8.

One story Dale might not remember is that at the time, the University was short of funds and upgrading the 360/67 was a problem.  We were supposed to order some sort of power generator for the V/6 ahead of time.  While Dale was off on vacation and I was Acting Director, a letter arrived from his boss (the Vice President, Academic) which I opened and read.  It said that we should not plan to install a new computer until the funding question was resolved.  I talked to Henry Ewaschko about this and we agreed to put the letter back in the envelop, reseal it and order the generator.  So the first step to installing the Amdahl was taken under a somewhat nefarious action.  When Dale returned and told us to delay we said it was too late.  Fortunately, the funding was found and the Amdahl installation went ahead.

Vic Yanda was quite a character and I'll leave it to Dale to tell you more about him.

Thanks again for continuing to enhance the MTS story, Jeff, and for coordinating all the material.

Best regards, Gerry Gabel

From: Dale Bent
Date: June 6, 2011 2:34:18 AM EDT
To: Gerry Gabel, Jeff Ogden
Subject: Re: Notes about MTS at the University of Alberta

Hi Gerry and Jeff

Gerry, we definitely have to get together to share more stories about those interesting times.

About Vic Yanda:   he was a senior systems analyst who specialized in the documentation and user support at the time we set up the Department of Computing Services.   Vic was a wild and wooly character who knew me from my frequent contacts with the computer centre and I think really wanted me to come over and straighten things out.   At least that what I think! After the new department was formed, Vic stayed with us for another ten
years or so before moving on.

Jeff, Gerry answered about our relationship with Amdahl.   We were the first Amdahl buy in Canada and an excellent reference account for them. We were often on the "tech tour" for new Amdahl customers.

Conversely, we went onto the shit list for IBM Canada.  Let me tell the story.

We desperately needed to upgrade from the IBM 360/67.  I went to my superior, an Associate VP, Academic, to get permission to go ahead.   He told me in no uncertain terms this was out of the question.   I went back
to our group with the bad news and received their advice.   We all agreed that we needed to get a much bigger computer and the Amdahl V/6 was the right choice.   Gerry stiffened up my spine and I tried again without success.   I was pretty heated about it and my boss told me if I didn't like it to go see the President.   So I did, and sold him that we needed to go the Amdahl route.  He was just superb.   The great service we were giving with MTS and the high satisfaction of senior chemists (the President has been a research chemist) had a lot to do with it.

This served us well when IBM got the news that we were going to single-source the purchase to Amdahl.  First, they invited me to a dinner at a golf club and ganged up on me.   The Branch Manager, Regional Vice-President, and lead U of A sales guy were there.  I told them that as far as I could determine that the Amdahl V/6 was a superior and more cost-effective machine, and we were getting a hell of a deal (we did). They questioned my judgement, my facts, and let me know that they intended to challenge my action.  And they did, contacting the President, my boss, the Director of Purchasing, and even members of the Board of Governors. But the order stuck.

Prior to this, the University of Alberta was always on a "tech tour" for IBM accounts, or for visiting IBM VPs.   I didn't see anyone above our account rep for five years. . . .

I did find out about Gerry and Henry Ewashecko "deep sixing" the order not to spend money for the power supply for the V/6.   This helps to explain, with the story above, why I was not on his [Vice President, Academic] favorite person list -- quite the opposite.  No matter -- we overwhelmed them with great computing!

Thanks again to Jeff for updating the history.    Please pass on our kind wishes to your colleagues.

With best wishes, Dale Bent.

From: John Stasiuk
Date: June 6, 2011 5:15:34 PM EDT
To: "Jeff Ogden"
Cc: "Gerry Gabel", "Dale Bent"
Subject: Re: Notes about MTS at the University of Alberta

Hi Jeff,
 
        I started with the Computing Center at the UofA in June 1963 as an Operator for the IBM 1620 Paper Tape System and we still ran the LGP 30 with paper tape as well at that time. The IBM 1620 was upgraded to Card Reader Punch and Printer and then to a tape system. A U of A developed 1620 Tape system called Go that allowed automatic job to job transition was implemented towards the end of the 1620. An IBM 7040 and IBM 1401 (to handle printing) replaced the 1620 system.
 
        Al Heyworth was the Computing Center Manager and Dr Don Scott was the Computing Center Director when the decision was made to order the IBM 360/67 (Al went to Medical Computing at the U of Toronto when all the goods weren't delivered for the IBM 360/67). The IBM 7040 system was becoming overloaded and IBM had been touting TSS. In theory TSS was going to be a very good system and the UofA needed something that would handle multiple users since computing was exploding and users were demanding more access.
 
       Al Heyworth, Russ James and myself were TSS trained by IBM at Poughkeepsie. We were not able to gen a TSS system that was stable for any real length of time so we had to use OS 360 when the IBM 360/67 arrived. We continued to try new TSS releases and fixes from IBM without ever ending up with what we thought would be a stable and useable system.
 
      Before OS 360 MVT was implemented, the Administrative Computing group got their own machine so that relieved a conflict between Administrative and Scientific computing. MVT may have been a solution but it was just to late.
 
      Just before going to MTS, we were running OS 360 MVT/HASP with the Student Oriented Batch Facility and DOS/ APL. I had made the HASP changes to allow SOB to run continuously and that really helped Student computing. APL was providing the time sharing portion that TSS never did.
 
      Gerry's group got MTS running in test mode and we were able to make the move to MTS.
 
      I think UM was supposed to go the TSS route as well, but when IBM didn't deliver a stable TSS they implemented MTS instead.
 
      Hope this sheds a little light on pre MTS at the UofA. Hope you can use some of the information above. Let me know.
 
     The Textform group was initially Grant Crawford, Dave Holberton and myself. Grant was the better programmer and had some good Computing Science training. We initially had many arguments about writing Textform in C and reverted to writing it in Assembler only because we did not believe we could do what we needed to do with C. That was a bad mistake because C+ and C++ would have handled the task. Textform was far superior to Word Perfect or Word and should easily have been the best  choice. But we needed to make it available to the non-MTS world and didn't do that.
 
         Anyway those are some of my thoughts.
 
take care,     John



Comments